Research work · 2003

Sonar signal processing based on the harmonic wavelet transform

Takaaki Musha, Tatsuya Kumazawa

IEEE Underwater Technology (UT)

Abstract

Conventionally a fast Fourier transform (FFT) algorithm is applied for digital beamforming of a sonar system, but this algorithm is inadequate for a sonar system with large bandwidth since the FFT can not be efficiently performed for wideband processing. To overcome this disadvantage of a fast Fourier transform algorithm, wideband processing by a wavelet transform is proposed. In this paper, we study the possibility of the harmonic wavelet transform proposed by Newland for applying to the wideband processing of a sonar system and it can be shown that the sonar signal processing including beamforming and replica correlation can be replaced by this wavelet algorithm instead of a FFT algorithm. A simulation for a planar array in a noisy environment is also conducted and it is confirmed that this method can improve the signal to noise ratio for wide band signals compared with the conventional FFT method.

DOI: 10.1109/ut.2002.1002441